Beth Milito, 41, drives home from work in her 2011 Sienna minivan. (Astrid Riecken/FOR THE WASHINGTON POST)

Beth Milito and her husband bought a 2011 Toyota Sienna based on friends’ recommendations and the minivan’s overall four-star safety rating to protect their four children.

But tucked into the details of the government’s crash test results was another rating that Milito said she never saw, which now has her wondering about her own safety. The front passenger seat on Milito’s Sienna received two out of five stars on the frontal crash test, a fall from the top five-star rating for that seat on the Sienna’s 2010 and older models.

The key difference: Starting with 2011 models, the federal government replaced an average-size male dummy with a smaller female dummy for some tests. When the 2011 Sienna was slammed into a barrier at 35 mph, the female dummy in the front passenger seat registered a 20 to 40 percent risk of being killed or seriously injured, according to the test data. The average for that class of vehicle is 15 percent.

“When we’re out and about as a family, I’m the one sitting in that seat,” said Milito, of Alexandria, after learning of the test results.

And she doesn’t know how the female dummy would fare behind the wheel, where she spends most of her car time commuting and ferrying kids. The star-rating system’s frontal crash test uses only the male dummy in the driver’s seat.

Consumer advocates say the female dummy’s subpar performance in some top-selling vehicles reveals a need to better study women and smaller people in collisions. Until recently, only male dummies were used during more than three decades of government testing aimed at helping car buyers choose between vehicles. The female dummy also mimics a 12-year-old child.

In general, experts say, the smaller the person, the fewer crash forces the body can tolerate. When cars wrap around trees or utility poles, for example, smaller drivers and passengers suffer more head, abdominal and pelvic injuries but fewer chest injuries than average-size people, according to the National Highway Traffic Safety Administration. Women’s less-muscular necks also make them more susceptible to whiplash, researchers say.

A 2011 study by the University of Virginia’s Center for Applied Biomechanics found that seat-belted female drivers in actual crashes had a 47 percent higher chance of serious injuries than belted male drivers in comparable collisions. For moderate injuries, that difference rose to 71 percent.

Auto safety watchdog groups say they’ve been pushing NHTSA to go beyond the average-sized male dummy since the agency launched the star-rating system in 1978. They say those tests should take into account not only women but the increasing elderly and obese populations and larger children who have outgrown child safety seats. The tests, they say, also miss average women who fall between the 50th percentile male dummy, which stands 5-feet-9 and 172 pounds, and the unusually petite female dummy, which is 4-feet-11 and 108 pounds.

The average American man is 5-feet-9 and 195 pounds, and the average American woman is 5-feet-4 and 165 pounds, according to Centers for Disease Control and Prevention statistics.

Child dummies, which are not included in the star-rating system, are used in separate tests of air bags and child safety seats.

“A lot of women do substantial [safety] research before going to buy a car,” said Joan Claybrook, a longtime consumer advocate who was head of NHTSA during the Carter administration, when the star-rating system was introduced. “Yet there’s not a whole lot of information about how cars impact women” in collisions.

Government data from police-reported crashes also show women are at greater risk of being hurt, particularly when they’re not behind the wheel. In the driver’s seat, men outnumber women by a ratio of 3 to 1 in vehicle fatalities. Men also drive 50 percent more than women — an average additional 5,000 miles annually.

While females comprise one-quarter of all driver fatalities, they make up half of all passengers killed, according to NHTSA. Because they’re on the road less, women are killed and injured at disproportionately higher rates than males, experts say.

The relative safety shortcomings in the front passenger seat for the female dummy went beyond the 2011 Sienna and its 2012 model, which improved to three stars for that seat. Other top-selling vehicles that earned three stars for that seat include the 2011 Honda CR-V, the 2012 Nissan Sentra and the 2012 Ford Fusion. The 2012 Acura TL received two stars for that seat.

NHTSA officials say they’ve used the female dummy in federal compliance crash tests since 2003, mostly to ensure air bags’ safety. Early versions of air bags severely injured or killed some smaller women and children.

However, those compliance tests gauge only whether a vehicle meets the government’s minimum safety standards. Consumers can’t readily see those results, and it’s the star-rating tests that reveal how one vehicle’s crash protections compare to another. The star ratings — which can be found at and appear on new car window stickers — use the car-buying market to encourage automakers to voluntarily exceed the minimum standards.

NHTSA officials said they recently included the female dummy in the frontal crash test as one of several changes to make the star-rating tests more rigorous. Even when not required in government tests, auto manufacturers have long used different dummies — including the female dummy and a larger 95th percentile male dummy — in internal tests.

The auto manufacturers had done so well adapting their vehicle designs to the previous tests — adding cushioning, reshaping arm rests — that most vehicles had achieved four and five stars, making it difficult for consumers to differentiate between them.

NHTSA officials warn that consumers shouldn’t compare crash-test ratings on vehicles in model years 2010 and earlier with those tested since because both the tests and how the results are calculated changed so significantly.

“Under the new test, a vehicle with three stars is not less safe than when it had five stars in the old program,” NHTSA spokeswoman Karen Aldana said. “The bar to get five stars has changed.”

Why a vehicle ended up with two or three stars vs. five is almost impossible for consumers to figure out. The test data posted on NHTSA’s Web site are so technical that consumers can’t tell whether the dummy registered a broken neck or a scratched knee.

Ronald Medford, NHTSA’s deputy administrator, said the star-rating tests are designed to complement the compliance testing, during which a female dummy is placed in the driver’s seat.

“You can be assured the protection for women is there in the federal standards,” he said.

In the separate star-rating tests that the public sees, Medford said, NHTSA now uses the female dummy in the driver’s seat for a 20-mph side crash test into a pole and behind the driver for the 38.5-mph test simulating a side impact at an intersection. The female dummy is not put in the driver’s seat on the frontal crash test, he said, because men drive more and die in greater numbers than female drivers.

Safety watchdog groups say leaving the female dummy out of the driver’s seat on the frontal tests leaves consumers with too little information, particularly in vehicles that appeal specifically to women, such as minivans. Frontal crashes account for 60 percent of all fatalities, according to NHTSA.

It’s more difficult to protect smaller drivers because they sit closer to the gas and brake pedals and more upright to see over the dashboard, consumer advocates and biomechanics experts say. That brings their heads and chests closer to the steering wheel and its air bag. The angle of their knees and hips as their shorter legs reach for the pedals also makes their legs more vulnerable, researchers say.

Heather McCann of Bethesda said she wants to know how the female dummy would do behind the wheel of her Honda CR-V, a model she usually sees women driving.

“My assumption going into it was that any car I bought would meet the minimum standards, or why else would it be on the road?” McCann said. “Then the five-star rating matters. Is it the gold standard [for safety] or is it not? It matters to me if something is three, four or five stars.”

Brian R. Lyons, a Toyota spokesman, said consumers shouldn’t compare the 2010 and 2011 Sienna star ratings because of the differences in the tests. The new Siennas aren’t less safe than previous designs, Lyons said; it’s just that the test is more stringent. Even so, he said, “We’re striving to achieve five stars for all our vehicles.This would be an area we’d be looking to enhance. . . . This vehicle getting a four [stars overall] means we’ll be going back and taking another look at it.

The nonprofit Insurance Institute for Highway Safety, which does its own crash tests, uses a male dummy driver for its frontal test and a female dummy driver in its side-impact test, said Raul Arbelaez, vice president of IIHS’s Vehicle Research Center.

Seat belts and air bags protect smaller people in frontal crashes in today’s vehicles better than they did 10 to 20 years ago, Arbelaez said. IIHS uses the female dummy in the side-impact test to ensure that newer side air bags fully protect the head, he said. Early side air bags didn’t cover the entire window, which left shorter drivers’ heads exposedbecause they were lower and farther forward than the average man.

When IIHS used a female dummy driver on limited research tests for frontal crashes, Arbelaez said, researchers saw no significant differences from the male dummy. “We know it [the male dummy] is working very well to reduce injuries in real-world crashes for all occupants, not just mid-size males,” he said.

Sheila Mitra-Sarkar, a research fellow in traffic safety at San Diego State University, said women’s injuries need further study because their vehicle use is generally markedly different from that of men.

One key reason women drive less is that they tend to choose jobs closer to home and family, she said. That helps protect them because they commute and run errands more on congested local roads, meaning their collisions occur at lower speeds.

However, the man typically drives when a couple heads out together, she said, leaving the woman in the front passenger seat, where she can get hit by an oncoming vehicle if the driver runs a red light. Not driving also may leave women less able to anticipate an impact, she said.

Even then, Mitra-Sarkar said, “We shouldn’t ignore the ones who are dying behind the wheel.”